There continues to be a need for effective yet safe treatments for anxiety and related disorders. Current treatments such as the benzodiazepines which target many of the gamma-aminobutyric acid receptor (GABA-A receptor) subtypes are effective, but patients suffer from a number of unwanted side-effects. The identification of compounds targeting only the anxiolytic alpha-2 GABA-A receptor subtype has been hampered by a lack of structure/function information needed for effective lead discovery and optimization. Phase I of this project will apply Neurion's NP2 technology and structure-based drug design platform to discover selective allosteric modulators for the alpha-2 GABA-A receptor. The phase I specific aims are: 1. We will generate initial structural knowledge about ligand-receptor interactions at the benzodiazepine (BZD) allosteric modulator site. We will use unnatural amino acid mutagenesis and electrophysiological assays to evaluate the binding interactions of a "modulator panel" of up to 50 previously described, structurally diverse compounds at six "probe positions" within the BZD site of both the alpha 2 & alpha 3 GABAAR subtypes. The modulator panel consists of known allosteric modulators that bind to the BZD site of the GABAAR and display varying degrees of functional subtype specificity. We will construct computational models of the GABAAR binding site that capture all this information. We will thus have experimentally validated computer models of drug-receptor interactions at both the alpha 2 & alpha 3 GABAAR subtypes. 2. We will use the computational models developed in Specific Aim 1 to computationally screen approximately 2 million commercially available compounds in addition to virtual combinatorial libraries and to select for purchase or to synthesize approximately 200 promising alpha 2 GABAAR subtype-selective molecules. We will electrophysiologically assay our 200 compounds for potency/efficacy and selectivity against alpha 1 and alpha 2 as well as alpha 3 and alpha 5 GABAAR subtypes. 3. We will choose the most promising approximately 30 leads from the compounds selected in Specific Aim 2, and conduct an additional round of unnatural mutagenesis and modeling as in Specific Aim 1 to further refine our models. 4. The refined model will be used to design and synthesize 10 novel compounds. We will assay the new compounds for potency/efficacy and selectivity against alphal, 2, 3, 5 GABAAR subtypes.